Research‎ > ‎

Algorithms and Software


 Sliding Organ RegistrationTraditional registration methods assume a smooth deformation field and enforce that smoothness via isotropic regularization.   We have devised a registration method that allows for sliding in the direction in the tangential direction at surfaces.
  • D. F. Pace, M. Niethammer, and S. R. Aylward, “Sliding geometries in deformable image registration,” in MICCAI’11: Proceedings of the Third international conference on Abdominal Imaging, Berlin, Heidelberg, 2011, pp. 141–148. 
  • Source code available in TubeTK
 Geometric MetamorphosisTraditional registration methods assume a single deformation field to account for differences between images.  However, if a pathology changes via infiltration or recession, then the pathology's deformation field may be locally discontinuous compared to the background deformation field which accounts for tissue displacement induced by the pathology.   Geometric Metamorphosis simultaneously solves for both deformation fields.
  • M. Niethammer, G. L. Hart, D. F. Pace, P. M. Vespa, A. Irimia, J. D. Van Horn, and S. R. Aylward, “Geometric metamorphosis,” in MICCAI’11: Proceedings of the 14th international conference on Medical image computing and computer-assisted intervention, Berlin, Heidelberg, 2011, pp. 639–646.
  • Source code available in CalaTK
 Vessel SegmentationThe extraction of the centerlines of tubular objects intwo and three-dimensional images is a part of many clinical image analysis tasks. One common approach to tubular object centerline extraction is based on intensity ridge traversal. We have developed a multi-scale traversal technique that is insensitive to noise, intensity and spatial variations common to vascular networks and a variety of imaging modalities.
  • S. R. Aylward and E. Bullitt, “Initialization, noise, singularities, and scale in height ridge traversal for tubular object centerline extraction,” Medical Imaging, IEEE Transactions on, vol. 21, no. 2, pp. 61–75, 2002. 
  • Source code available in TubeTK
 Vessel-based RegistrationOur method aligns a source image with a target image by registering a model of the tubes in the source image directly with the target image. Time can be spent to extract an accurate model of the tubes in the source image. Multiple target images can then be registered with that model without additional extractions. Our registration method builds upon the principles of our tubular object segmentation work that combines dynamic-scale central ridge traversal with radius estimation. In particular, our registration method’s consistency stems from incorporating multi-scale ridge and radius measures into the model-image match metric. Additionally, the method’s speed is due in part to the use of coarse-to-fine optimization strategies that are enabled by measures made during model extraction and by the parameters inherent to the model-image match metric.
  • S. Aylward, J. Jomier, S. Weeks, and E. Bullitt, “Registration and analysis of vascular images,” INTERNATIONAL JOURNAL OF COMPUTER VISION, vol. 55, no. 2–3, pp. 123–138, Dec. 2003. 
  • Source code available in TubeTK
 Spatial Graphs: Vascular Network CharacterizationOur interest in characterizing intra-cranial vasculature arises from the mounting evidence that a genetic relationship exists between mental disorders and vascular network formation. It has been established that during development vascular endothelial growth factors not only spur vessel and tissue growth but also direct tissue differentiation and the formation of organs.

Graph methods that summarize vasculature by its branching topology are not sufficient for the statistical characterization of a population of intra-cranial vascular networks. Intra-cranial vascular networks are typified by topological variations and long, wandering paths between branch points. We have developed a graph-based representation, called spatial graphs, that captures both the branching patterns and the spatial locations of vascular networks. Furthermore, we have developed companion methods that allow spatial graphs to (1) statistically characterize populations of vascular networks, (2) generate the central vascular net- work of a population of vascular networks, and (3) distinguish between populations of vascular networks.
  • S. Aylward, J. Jomier, C. Vivert, V. LeDigarcher, and E. Bullitt, “Spatial graphs for intra-cranial vascular network characterization, generation, and discrimination,” in MEDICAL IMAGE COMPUTING AND COMPUTER-ASSISTED INTERVENTION - MICCAI 2005, PT 1, 2005, vol. 3749, pp. 59–66. 
  • Source code to be released with TubeTK (in the future)
 Vessel Tortuosity Characterization
The clinical recognition of abnormal vasculartortuosity, or excessive bending, twisting, and winding, isimportant to the diagnosis of many diseases. Automated detectionand quantitation of abnormal vascular tortuosity from three-dimensional (3D) medical image data would therefore be of value.
  • E. Bullitt, D. Zeng, B. Mortamet, A. Ghosh, S. R. Aylward, W. Lin, B. L. Marks, and K. Smith, “The effects of healthy aging on intracerebral blood vessels visualized by magnetic resonance angiography,” NEUROBIOLOGY OF AGING, vol. 31, no. 2, pp. 290–300, Feb. 2010. 
  • Source code to be released with TubeTK (in the future)

Kitware Software








External Projects

 Python: SciPy, NumPy, PyPipe,